|Home | About | Journals | Submit | Contact Us | Français|
The bile salt export pump (BSEP, ABCB11) is the major determinant of bile salt dependent bile secretion and its deficiency leads to cholestatic liver injury. BSEP/Bsep gene expression is regulated by the nuclear farnesoid X receptor (FXR). However, BSEP expression is retained in the liver of the Fxr−/− mice although reduced, indicating that there may be additional transcriptional factors that regulate its expression. The NF-E2-related factor-2 (Nrf2) plays a major role in response to oxidative stress by binding to antioxidant-responsive elements (ARE) that regulate many hepatic Phase I and II enzymes as well as hepatic efflux transporters. Computer software analysis of human BSEP reveals two Maf recognition elements (MAREs) from the sequence in the proximal promoter region where Nrf2 may bind. In this study we examine if Nrf2 plays a role in human BSEP expression and whether this might be mediated by the MAREs. Oltipraz, a potent activator of Nrf2, increased BSEP mRNA expression by ~ 7-fold in HepG2 cells and protein by ~ 70% in human hepatocytes. siRNAs lowered NRF2 expression in HepG2 cells and prevented the up-regulation of BSEP by oltipraz. Human BSEP promoter activity was stimulated by Nrf2 in a dose-dependent manner in luciferase reporter assays. Mutations of the predicted MARE1, but not MARE2, abolished this Nrf2 transcriptional activation. ChIP assays also demonstrated that Nrf2 specifically bound to MARE1, but not MARE2 regions in the BSEP promoter in HepG2 cells. Electrophoretic mobility shift assays further demonstrated direct binding of MARE1 in the BSEP promoter.
Nrf2 is a positive transcriptional regulator of human BSEP expression. Pharmacological activation of Nrf2 may be beneficial for cholestatic liver injury.
The bile salt export pump (BSEP, ABCB11) is a member of the ATP-binding cassette (ABC) superfamily of transporters(1). It is primarily expressed in the liver where it localizes to the canalicular membrane of hepatocytes. BSEP/Bsep is the major determinant of bile salt-dependent bile secretion, and secretes monovalent conjugated bile acids from hepatocytes(2). Genetic deficiencies of BSEP lead to progressive cholestatic liver injury(3, 4) and are risk factors for hepatocellular carcinoma(5).
BSEP/Bsep expression is highly regulated by the nuclear farnesoid X receptor (FXR, NR1H4) which heterodimerizes with the retinoid X receptor (RXR, NR2B), and binds to an inverted repeat (IR)-1 element in the BSEP promoter(6). BSEP/Bsep is also regulated by the liver receptor homolog-1 (LRH-1) and its expression is decreased in hepatocyte-specific Lrh−/− mice(7). However, Bsep expression is still preserved in both Fxr−/− and Lrh−/− mice(7–9), suggesting that there may be additional transcriptional factors that regulate its expression.
The transcription factor NF-E2-related factor-2 (Nrf2) plays an important role in maintaining Redox homeostasis by regulating the expression of many Phase I and Phase II drug-metabolizing and detoxification enzymes, such as NAD(P)H quinone oxidoreductase (Nqo1), and glutathione-S-transferase (GST)(10). In addition, Nrf2 also regulates a few Phase III transport proteins, including the multidrug resistance-associated proteins (Mrp) 1–4(11).
Nrf2 is a 605 residue ubiquitous protein that belongs to the small family of basic leucine zipper transcription factors(12, 13). Activation of Nrf2 is controlled by the actin-associated kelch-domain protein 1 (Keap1), which acts as a negative regulator of Nrf2(14). During basal conditions, Keap1 binds to Nrf2 and sequesters Nrf2 cytoplasmically(15). Under oxidative stress, Nrf2 dissociates from Keap1 and translocates to the nucleus, where it heterodimerizes with the small musculo-aponeurotic fibrosacroma (Maf) protein. This complex then binds to an antioxidant responsive element (ARE) or an AP1-NF-E2 tandem repeat sequence, called Maf recognition elements (MARE), initiating transcription(12, 16). Computer analysis reveals that human BSEP contains two MARE sites in its proximal promoter region, suggesting that Nrf2 may regulate BSEP expression.
In the present study, we found that Nrf2 activators, in particular oltipraz (OPZ), induced BSEP mRNA and protein expression in HepG2 cells and human hepatocytes. Knockdown of Nrf2 by siRNA diminished the increase in BSEP mRNA expression. Human BSEP promoter reporter assays demonstrated that one of the computer-predicted MAREs mediated Nrf2 regulation in promoter activity. The involvement of Nrf2 in regulation of BSEP expression was further confirmed by ChIP and gel mobility shift assays in HepG2 cells. These findings indicate that Nrf2 is a positive regulator of BSEP expression, suggesting activation of Nrf2 might be of benefit in cholestatic liver injury.
Oltipraz was purchased from Axxora (San Diego, CA). Other chemicals were obtained from Sigma-Aldrich (St. Louis, MO). Polyclonal antibody against human BSEP was purchased from Kamiya Biomedical (Seattle, WA). Polyclonal antibody against human Nrf2 (sc-722) and siRNA-Nrf2 were obtained from Santa Cruz Biotechnologies (Santa Cruz, CA). siRNA-control (a non-targeting siRNA) was purchased from Ambion (Foster, CA). A dual-luciferase assay kit was purchased from Promega (Madison, WI).
The HepG2 cell line was obtained from the American Type Culture Collection (Manassas, VA) and cultured in Eagle's minimal essential medium (EMEM) supplemented with 10% fetal calf serum (Invitrogen; Carlsbad, CA), 100 units/ml penicillin G, and 100 µg/ml streptomycin. Primary human hepatocytes were obtained from the Liver Tissue Procurement and Distribution System of the National Institutes of Health (Dr. Stephen Strom; University of Pittsburgh, PA) and cultured as described previously(17).
A 2.5 kb DNA fragment containing the human BSEP proximal promoter region was amplified from human genomic DNA (Clontech Laboratories; Mountain View, CA) by PCR using primers (Table 1) targeting nucleotide −2563 bp and +4 bp from the transcription start site (from GenBank Accession No. AF091582). The sequence was verified by DNA sequencing and cloned into pGL3-Basic vector (Promega), named p-2563/+4-Luc. Subsequently, a shorter fragment containing −596 to +4 bp of human BSEP promoter was also cloned into pGL3-Basic vector, named p-596/+4-Luc. A full-length mouse Nrf2 expression plasmid was kindly provided by Dr. Jefferson Chan, (University of California; Irvine, California). MatInspector program (www.genomatix.de) was used to analyze the human BSEP promoter sequence, and two putative MAREs were identified. To mutate the MAREs, the QuikChange™ site-directed mutagenesis kit (Stratagene; Cedar creek, TX) was used according to the manufacturer's instructions (primers listed in Table 1). All reporter constructs were verified by DNA sequencing at the W.M. Keck Biotechnology Resource Laboratory at Yale University.
Total RNA was isolated from HepG2 cells and hepatocytes using Trizol reagent (Invitrogen) and purified using Qiagen RNA clean-up kit. cDNA was synthesized using AffinityScript Multiple Temperature cDNA synthesis kit (Stratagene). Real-time PCR was performed in an ABI 7500 SDS (Applied Biosystems; Foster city, CA). The level of human BSEP, Nrf2, and GAPDH mRNA were measured using TaqMan gene expression Assays ID: Hs00184824_m1, Hs00232352_m1, and ID: Hs99999905_m1, respectively (Applied Biosystems).
HepG2 cells were plated in 24-well plates. Twenty-four hours after seeding, cells were transiently transfected with 1.8 µl of FuGENE HD (Roche; Indianapolis, IN) together with 0.6 µg of total plasmid DNA per well consisting of 0.3 µg promoter reporter construct or empty pGL3-Basic vector, 2.5 – 50 ng Nrf2 expression plasmid, 10 ng phRL-CMV (renilla luciferase plasmid), and 0.24 – 0.29 µg of pcDNA 3.1. After transfection for 24 h, cells were treated for 24 h with 0.05% DMSO or oltipraz (50 µM). The firefly and renilla luciferase activities were determined using a Dual-Luciferase Reporter Assay (Promega) in a Synergy2 microplate reader (BioTek; Winooski, VT). The promoter activity was calculated by normalizing the firefly luminescence to the renilla luminescence signal and the ratio of promoter construct over control is presented. For knocking down Nrf2, overnight-seeded HepG2 cells were transfected with 150 pmols Nrf2 siRNA or control (non-targeting) siRNA with lipofectamine 2000 (Invitrogen) before treatment with either DMSO or OPZ (50 µM).
Cells were lysed with RIPA buffer (50 mM Tris·HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing Halt protease inhibitor cocktail (Pierce; Rockford, IL). The protein concentration was determined using BCA protein assay (Pierce). Proteins were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using a 7.5 % Tricine gel, electrophoretically transferred to nitrocellulose membranes and assayed as previously described(18). The blot was quantified and analyzed using FOTODYNE Imaging and TL-100 image analysis software.
HepG2 cells were grown on glass coverslips, transfected either with pcDNA 3.1 or Nrf2 expression plasmid, and treated either with DMSO or oltipraz (50 µM) for 24 h. Cells were fixed with 4% paraformaldehyde for 30 min and incubated with anti-Nrf2 polyclonal antibody (1:25) for 1 h at room temperature. Alexa488 anti-rabbit immunoglobulin G (Molecular Probes; Eugene, OR) was used as the secondary antibody, and cells were incubated with the TOPRO-3 (Molecular Probes) for nuclear staining. The fluorescent image was visualized with a Zeiss LSM510 confocal scanning microscope (Carl Zeiss Inc.; Thornwood, NY), and images were processed with Adobe PhotoshopCS (Moutainview, CA).
ChIP analysis was performed using the ChIP assay kit, according to the manufacturer’s instruction (Millipore-Upstate; Temecula, CA). Briefly, HepG2 cells were grown in 100-mm dishes to 80% confluence. The cells were treated with either DMSO or oltipraz (50 µM) for 48 h before harvesting. Chromatin samples were digested with nuclease into length averaging between 150–900 bp. Prior to immunoprecipitation, chromatin samples (input) were precleared using protein G beads. The precleared chromatin samples were then incubated with anti-Nrf2 polyclonal antibody or rabbit IgG as a negative control. RNA Polymerase II antibody and primers specific for GAPDH assay were used as a positive control. After DNA purification, PCR was performed using 3 µl of the DNA extracted from the input for each immunoprecipitation reaction. Thermal cycling profile for PCR amplification was set as manufacturer’s instruction (primers listed in Table 1). PCR reaction was electrophoresed through 2.5% agarose gel.
Nuclear protein extract from HepG2 cells was isolated by NE-PER kit (Pierce). EMSA was performed using DIG-gel shift Kit, 2nd Generation (Roche) as the manufacturer’s instruction. The sequences of probes were listed in Table 1. In brief, 5 ug of nuclear protein was incubated in the binding solution containing 15 pmol of digoxigenin-labeled probe. A 100-fold excess of unlabeled probe was added to the reaction for competition assays.
Data are presented as the mean ± standard error of the mean (S.E.M.), and n values represent sample size. Differences between groups were determined using Student’s t-test or One-way analysis of variance followed by Dunnett’s test (SigmaStat program, Jandel Scientific). Differences with p<0.05 were considered statistically significant.
To determine whether Nrf2 is involved in the expression of BSEP, we treated HepG2 cells with classical Nrf2 activators, including OPZ, t-BHQ and quercetin. CDCA and 4-PBA, known inducers for human BSEP, were used as positive controls(6, 19). As shown in Figure 1A, OPZ, quercetin, and t-BHQ significantly increased BSEP mRNA expression by ~ 7X, 5X, and 3X, respectively after 48 h treatment. In contrast, mRNA expression of another bile salt transporter, the organic solute transporter alpha (Ostα), is not changed by the same treatment (data not shown). The induction of BSEP expression by OPZ was dose-dependent with an optimal concentration of 50 µM (Fig. 1B), as well as time dependent (Fig. 1C). BSEP mRNA expression was significantly induced by 24 h and maximal induction was seen at 48 h and remained sustained 72 h after treatment. In human hepatocytes, the addition of OPZ also led to increased BSEP message levels. However, the onset of elevation in BSEP mRNA levels occurred more rapidly than in HepG2 cells. This induction was seen starting from 3 h and gradually declined by 24 h (supplementary Fig. 1S). Taken together, these results indicate that Nrf2 agonists can induce BSEP gene transcription and that Nrf2 is likely involved.
The basal expression of BSEP protein is very low in HepG2 cells and is undetectable by Western blot using currently available antibodies (unpublished data). Therefore we examined the expression of BSEP protein in human hepatocytes. After 48 h treatment OPZ significantly increased BSEP protein expression in human hepatocytes by 1.7-fold compared to DMSO treated controls, while the FXR agonist CDCA (50 µM) increased BSEP protein 4.2-fold (Fig. 1D). Based upon time-dependence of OPZ induction in human hepatocytes, there seems to be a discrepancy between OPZ-induced mRNA (maximum after 3 h) and protein expression (maximum after 48 h). These results suggest the possibility that OPZ may also have post-translational effects on BSEP expression.
HepG2 cells were treated with DMSO or OPZ and total protein, cytosolic and nuclear extracts were assessed by immunoblot analysis. The total Nrf2 protein was increased in OPZ-treated compared to DMSO control cells (Fig. 2A). Nrf2 accumulated in the nucleus in both conditions and again showed the large increase in the amount of protein after OPZ treatment. Nrf2 was not detectable in the cytosolic fraction, which is likely due to Nrf2 activators which stabilize Nrf2 in the nucleus, as recently described(20). This stabilization in the nucleus may be mediated either by preventing or reducing access of Keap1 to Nrf2 molecules for degradation in the cytoplasm. Immunofluorescence studies in HepG2 cells and human hepatocytes also confirmed these immunoblot data (supplementary Fig. 2S)
To investigate if Nrf2 directly regulates BSEP mRNA expression, we first analyzed a 3 kb sequence of the human BSEP promoter region using MatInspector. Two putative MAREs were identified in the proximal promoter (Fig. 3). To test if these MAREs regulate BSEP expression, we co-transfected Nrf2 expression plasmids and BSEP promoter constructs into HepG2 cells. As shown in Figure 4, both p-596/+4-Luc and p-2563/+4-Luc resulted in only minimal increases in luciferase activity compared with the empty pGL3-Basic vector (3.9-fold, and 1.2-fold, respectively). In contrast, promoter activity was significantly increased in a dose-dependent manner when Nrf2 was co-transfected. Interestingly, the p-594/+4-Luc construct demonstrated higher inducibility (34 to 97-fold) than the p-2563/+4-Luc construct (14 to 47-fold) when the same amount of Nrf2 was co-transfected (Fig. 4).
In transfected HepG2 cells, additions of OPZ did not result in further increases in luciferase activity. As can be seen in Fig. 2B, the transfected Nrf2 was already translocated into nuclei. This is most likely due to insufficient amounts of Keap1 necessary to retain the exogenously expressed Nrf2 in the cytosol.
To examine the relative functional importance of these two MAREs in the BSEP promoter, we mutated the core region of MARE1 in p-596/+4-Luc as well as MARE1 and MARE2 in p-2563/+4-Luc. Both wild-type and mutant constructs were then co-transfected with different amounts of the Nrf2 expression plasmid. Mutation of MARE2 had no significant effect on basal or Nrf2-induced transcriptional activity (Fig. 5A). In contrast, mutation of MARE1 completely eliminated both basal and Nrf2-inducible promoter activity both in the p-596/+4-Luc (data not shown) and in p-2563/+4-Luc constructs (Fig. 5B). These results indicate that Nrf2 activation is mediated by the MARE1 at −175 to −195 bp in the BSEP promoter.
To further confirm if Nrf2 binds specifically to these MAREs in the BSEP promoter, a ChIP assay was employed using chromatins prepared from HepG2 cells treated with DMSO or OPZ and probed with oligonucleotides flanking the MARE sites. As shown in Figure 6A, a band was seen in all inputs, but only chromatins treated with OPZ and immunoprecipitated with Nrf2 antibodies showed a strong band targeting the MARE1 region. In contrast, no PCR products were detected after targeting the MARE2 region when chromatins immunoprecipitated with Nrf2 were used as template (Fig. 6B). These findings suggest that Nrf2 was specifically recruited and bound to the MARE1 motif in the proximal BSEP promoter.
To verify the direct binding of MARE1 to the BSEP promoter, EMSA was performed with an oligonucleotide probe corresponding to the MARE1-motif found in the BSEP promoter (Fig. 6C). When the probe was incubated with nuclear protein extracted from HepG2 cells, a band was observed (lane 2). Formation of the band was completely abolished by adding a 100-fold excess of unlabeled competitor corresponding to the sequence for the MARE1-motif (lane 3) or a known ARE element from NAD(P)H quinone oxidoreductase (Nqo1) (lane 5). However, the band was not affected by the addition of mutated competitor (lane 4). These results strongly indicate that MARE1 is associated with Nrf2 in the human BSEP promoter.
Finally, to assess if Nrf2 plays a functional role in BSEP expression, we knocked down Nrf2 expression using siRNA in HepG2 cells. We were able to significantly reduce Nrf2 mRNA expression by approximately 35% 72 h after transfection, and reduced Nrf2 protein expression was also observed (Fig. 7A, B). BSEP mRNA expression was decreased by 35% in Nrf2 knocked-down cells (Fig. 7C). Additions of OPZ for 48 h in control siRNA cells significantly induced BSEP mRNA expression by 1.7-fold compared to DMSO. In contrast, BSEP mRNA expression was not significantly changed in Nrf2 knocked-down cells. The induction of BSEP mRNA expression by OPZ in control siRNA cells was lower than in non-transfected cells illustrated previously in Fig. 1. This could be due to the non-specific side effects of RNAi transfection. Taken together, this result indicates a specific role for Nrf2 in mediating BSEP basal expression and induction by OPZ in HepG2 cells.
Our current study indicates that Nrf2 can positively regulate human BSEP gene expression via binding to an MARE element at −175 to −195 bp in its promoter. This conclusion is supported by the following results. 1) Nrf2 activators induced both BSEP mRNA or protein expression in HepG2 cells and human hepatocytes. In particular, OPZ was the most potent agonist for inducing BSEP mRNA expression among the tested compounds. 2) Nrf2 accumulated in the nucleus in human hepatocytes and HepG2 cells when exposed to OPZ. 3) Reporter assays demonstrated that Nrf2 activated the BSEP proximal promoter in a dose-dependent manner, and mutation of MARE1 but not MARE2 abolished this activation. 4) ChIP assays further confirmed that Nrf2 binds to the genomic BSEP promoter locus through MARE1 but not MARE2. 5) Electrophoretic mobility shift assays substantiated direct binding of MARE1 in the BSEP promoter. Finally, 6) Lowering Nrf2 expression by siRNA in HepG2 cells prevented the up-regulation of BSEP by OPZ treatment.
BSEP is the major determinant in bile salt-dependent bile formation in the liver. The functional importance of human BSEP is emphasized by genetic deficiencies of this gene that result in cholestatic liver diseases, including progressive familial intrahepatic cholestasis type 2 (PFIC2), benign recurrent intrahepatic cholestasis type 2 (BRIC2), and intrahepatic cholestasis of pregnancy (ICP)(21). Previous studies have established that the nuclear receptors FXR and LRH-1 regulate the human BSEP promoter via their specific response elements(6, 22). The functional roles of these two nuclear receptors in Bsep expression is supported by studies in knockout mouse models where Bsep mRNA expression is reduced(7, 8). However, the persistent expression of Bsep in these animals suggests that there should be additional transcription factor(s) that regulate BSEP/Bsep expression. Since Nrf2 regulates the expression of other ABC transporters, we reasoned that Nrf2 might also regulate human BSEP promoter activity. Although in-silico methods predicted two MAREs in the BSEP promoter, our results demonstrated that only one of them (i.e. MARE1) is functional. Motohashi and coworkers report that the nuclear transcription factor AP-1 can also bind to the consensus MARE (TGCTGAC(G)TCAGCA) sequence(16). Whether AP-1 binds to the MARE1 in the BSEP promoter remains to be determined. To check if the identified MARE1 is conserved in rodents, we aligned the genomic sequences of human, rat and mouse BSEP/Bsep promoter regions. Interestingly, the core MARE1 is conserved in human and rat, but not in mouse. MatInspector software also did not find MAREs in mouse Bsep. Taken together, these analyses imply that Nrf2 may not regulate mouse Bsep as suggested by other studies(23).
The functional role of Nrf2 in cytodetoxification is well-established as it directly regulates many Phase I and II enzymes by binding to the ARE/MARE sites in their promoters. Recent studies also indicate that Nrf2 plays an important role in regulating phase III transporters: including human MRP4 (our unpublished data) and Mrp1-4 in rodents(11). ChIP assays using Nrf2 antibody suggest there are AREs in the Mrp1-4 promoters(11). Nrf2 is required for both constitutive and inducible expression of Mrp1 in fibroblasts(24), while Mrp2-4 are up-regulated following treatment with Nrf2 activators(11). In addition, Mdr1 is increased in human hepatocytes treated with the Nrf2 activator, OPZ(25). Taken together with the results of this present study, agents that up-regulate Nrf2 activity might be of therapeutic benefit for cholestatic liver diseases.
Indeed, a recent study suggests ursodeoxycholic acid stimulates Nrf2-mediated up-regulation of Mrp2, 3 and 4 in mice via AREs in their promoters(23). However, Bsep mRNA expression was not significantly changed in that study. In agreement with this report, we also did not detect significant changes of Bsep mRNA expression in Nrf2−/− mice when compared to wild-type (unpublished data). Since the MARE1 sequence we have identified in human BSEP is not conserved in the mouse, this may explain why prior studies using Keap1 knockdown and Nrf2−/− mice found no effects on Bsep expression in this species(23). Thus, factors other than Nrf2 likely maintain expression of Bsep in Fxr−/− and Lrh−/− null mice.
During cholestasis, elevated concentrations of bile acids in hepatocytes activate FXR/RXR, which increases BSEP expression. In addition, cholestasis results in increased production of reactive oxygen species (ROS) and oxidative stress(26). Recent evidence indicates cytotoxic bile acids, including LCA, CDCA and DCA, are capable of activating nuclear Nrf2 and lead to induction of cytoprotective genes that help protect cells against bile acid toxicity(26). Thus, based on the results of this study, it is possible that Nrf2/MARE may work in concert with FXR/RXR to up-regulate human BSEP gene expression in an attempt to accelerate removal of bile acids from the liver as part of the adaptive response to cholestatic liver injury(27).
In summary, this study provides evidence that oltipraz, an Nrf2 activator, can function as another positive transcriptional regulator of the human BSEP promoter. We have identified that its induction is regulated via the Nrf2/MARE pathway. These findings not only expand our current knowledge of the regulatory mechanism of the human BSEP gene but may also provide a novel pharmacological approach for drug development and future therapeutic strategies for cholestatic liver disease. Most recently, the protective effect of oltipraz against intrahepatic cholestasis induced by ANIT has been established in an animal model as well(28).
We thank Albert Mennone for technical assistance and Dr. Thomas Kensler (Johns Hopkins University, Maryland) for providing liver tissue samples from Nrf2 knockout mice.
*This work was supported by National Institutes of Health Grant R37 DK 25636 (to JLB) and P30 DK34989. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.